Death of a PSU

[Joseph Noci 1Participant @josephnoci1]

A 22volt zener is appropriate – As Robert said – it catches spikes on the chip supply line and will regulate or clamp them to 22V. Chances are that if the zener does not measure short circuit it will be ok. The output voltage is controlled by the Optocoupler and the TL431 in the circuit – the latter may differ from your PSU.

Note that I did not say the reference circuit I gave is the actual circuit – it is a REFERENCE circuit provided by the chip manufacturer to aid the designer – Robin's psu will be a close derivative.

The part number MBRF2045 is a schottky rectifier, so most likely the output rectifier ( place of D3 in the circuit), I am sure that device will be operational still.

I do go along with the general persuasion that you just ditch it and get a replacement, but if you are like me and cannot leave things alone…go ahead and try to fix it!

It's not magic and the important thing is to be safe in the process. If you have no experience in the subject, and wish to have a go at replacing parts and seeing what happens, ok, just beware of high voltages on that circuit.

WRT the reference circuit, EVEYTHING to the left of the transformer and optocoupler is a life-taking high voltage. There is NO safe earth or ground reference in that part of the circuit, so you cannot easily connect oscilloscope probe grounds, etc to measure anything.

This danger is easily seen in the circuit, but is magnified on your actual board, as you cannot draw a nice visual line separating the danger zones. High voltage can be present anywhere. Do not handle the board at all in a manner that may allow skin contact with any pcb conductor or component lead. Make sure that heatsink is actually grounded, else don't touch it. Don't touch the PSU with mains on! And when the mains is off and disconnected, BEWARE of the high voltage still remaining in the region of the input filter electrolytic capacitors – 'C3' in the circuit…

Just be careful and attentive..

Joe

[Maurice TaylorParticipant @mauricetaylor82093]

When working on this type of power supply would it be safer to power it with an isolating transformer,like we used to do with AC/DC radios and tv’s ?

[Stuart BridgerParticipant @stuartbridger82290]

I'm with Robert on this one SMPSU's aren't easy to troubleshoot and repair.

[Robert Atkinson 2Participant @robertatkinson2]

While an isolating transformer will give protection from shocks to other grounds it won't help with the 350V DC on the PCB. Unless the whole workshop is isolated conneing test equipment like an oscilloscope removes the isolation.

Be carefull out there!

Robert G8RPI

[Robin GrahamParticipant @robingraham42208]

Thank you for the warning Robert . There was actually quite a lot hidden under the the conditional 'if' in my last post. Such as (inter alia) 'if ever I feel I understand this well enough to proceed safely" – which, at the moment, I don't, as you have rightly observed. Doesn't mean I can't learn by asking naive questions though!

Stay safe, Robin.

[Joseph Noci 1Participant @josephnoci1]

Posted by Maurice Taylor on 26/07/2020 21:52:31:

When working on this type of power supply would it be safer to power it with an isolating transformer,like we used to do with AC/DC radios and tv’s ?

Yes Maurice, you can do that – exactly how it IS done in development labs and production test.

You don't need isolation on the 'isolated' side of the isolation transformer – that is what it is there for – so with a proper isolation transformer in place, you can then ground ( to mains ground/earth) the negative lead of the rectified PSU 'mains' input side. You can then also connect your oscilloscope ground lead to that same ground, safely, and probe about. HOWEVER .. the high voltage still exists as before, nothing has changed, you only have a safe ground reference.

That's why I said – There is NO safe earth or ground reference in that part of the circuit, so you cannot easily connect oscilloscope probe grounds, etc to measure anything-

'you cannot easily connect …' you need an isolation transformer and some Care..

That supply is so simple – no switching FET, etc, and generally they pop the controller – The voltage reference optocoupler and TL431 reference normally are well isolated and do not suffer damage. So it is not as daunting a repair as may seem – check the zener, check the snubber diode ( the diode on the PCB top side) , just above the burnt underside, and replace any damaged, as well as the control IC. If it does not work, or pops loudly again, ditch it…

Joe

[John HaineParticipant @johnhaine32865]

As Robin has said he will use the other PSU he has or maybe fit a new one, maybe we could give the defunct unit a decent burial and move on?

[Joseph Noci 1Participant @josephnoci1]

Don't know – ask Robin?

[Grizzly bearParticipant @grizzlybear]

I haven't read all the replies to this post, this maybe useful or not.

Seven pin I.C's (Offline regulators) are used in washing machines to battery chargers for drills etc.

It is not uncommon for them to fail.

**LINK**

Good luck, Bear..

[Robin GrahamParticipant @robingraham42208]

Posted by Joseph Noci 1 on 27/07/2020 12:19:31:

Don't know – ask Robin?

Well, my answer to John if he were to ask me would be move on, don't read, that's up to you as an individual.

For me personally this has been a very interesting discussion – for the first time in my life I think I am on the brink of the threshold of understanding an electronic circuit. I could have just bought a replacement PSU – it wouldn't have left me starving. But I wanted to understand what had happened and if I can repair it it would give me a buzz, of the life-enhancing rather than lethal type I would hope!

Robin.

 

 

Edited By Robin Graham on 28/07/2020 02:59:04

Edited By Robin Graham on 28/07/2020 03:00:48

Edited By Robin Graham on 28/07/2020 03:01:28

[Stuart BridgerParticipant @stuartbridger82290]

Switched Mode Power Supplies are a technology that in theory are very simple, but in practice very tricky to get right. By right I mean stable and reliable. Much improved over the years as many devices have been by introduction of integrated control ICs. When I was a lad, most were controlled by discrete components and would go bang very frequently and often for no apparent reason. I used to work on early CAD terminals, which had a PSU where its main switching device device would fail short cicuit often. We had a big stock of replacements. TIPL755A, I can still remember the number after 35 years….

[Author]

Posts